DIVISIONS OF CONCEPTS AND TERMS. 51 



less variable. This variation, however, unlike that which attaches 

 to subjective intension, is only accidental. And arising as it does 

 from inherent and inevitable defects of clearness in language and 

 thought, it need not prevent us from demanding that as far as 

 possible the connotation of our concepts and terms be kept un 

 changed throughout our reasoning processes. 



We may next inquire on what principles the attributes that 

 form the connotation of a term are selected. What are to be the 

 limits of the group? There has been a great deal of controversy 

 and not a little confusion as to whether connotation should in 

 clude all the attributes, known and unknown, actually common to 

 all the things denoted by the term or all known to each indivi 

 dual mind to be common or all known by men generally to be 

 common or only some of the latter, namely, those fixed upon 

 by competent authorities and understood to be the test of classi 

 fication in the sense that no individual thing will be recognized as 

 a member of the class in question without possessing all of them. 

 But from the distinctions we have indicated between content, 

 connotation and comprehension, it will be seen that the last 

 alternative just mentioned gives the connotation. The attributes 

 included in this are, of course, those most directly implied by the 

 name, those understood to be most fimdamental and important in 

 the nature of that class of objects ; not those that are only in 

 directly implied, in the sense that they are properties necessarily 

 connected with, and derivable from, the more fundamental ones 

 directly implied ; nor those that may be known by some people 

 or by all people to be actually common to all members of the 

 class, though having no apparent connexion with the funda 

 mental attributes, and which may be merely suggested by the 

 mention of the name through some subjective association of ideas. 



Thus, for example, we take the connotation of equilateral triangle to be 

 &quot; plane, rectilinear figure with three equal sides &quot; ; we do not include the pro 

 perty of being &quot;equiangular,&quot; nor any other of the innumerable properties 

 brought to light by the geometrical study of such triangles. We take the 

 connotation of man to be &quot; rational animal &quot; ; we do not include the attributes 

 of &quot; cooking his food &quot; or &quot; using tools,&quot; though these are properties that 

 follow from his rationality ; nor the attribute of &quot; two-legged,&quot; though this is 

 de jacto naturally common to all men. Similarly, though all known &quot; rumi 

 nants &quot; are * cloven-hoofed,&quot; we do not regard this attribute as essential to 

 the ruminant, nor include it in the connotation of ruminant ; nor the attri 

 bute of &quot; Australian &quot; in the connotation of &quot;kangaroo &quot;. And although the 

 notion of &quot; wearing a soutane &quot; is inseparably associated in our mind with 



4* 



